Imprint

Philosophers’

B 
 volume 13, no. 2 ertrand Russell is often credited with bringing the sig-
 january 2013 nificance of relations to philosophers’ attention. He not only
recognizes relations as interesting in their own right but is
convinced that one’s account of relations bears on many other cen-
tral philosophical issues:

Suárez on the Reduction

The question of relations is one of the most important
that arise in philosophy, as most other issues turn on it:
monism and pluralism; the question whether anything is
true except the whole of truth, or wholly real except the

of Categorical Relations
whole of reality; idealism and realism, in some of their
forms; perhaps the whole existence of philosophy as a
subject distinct from science and possessing a method of
its own.1

It is certainly the case that much of the attention analytic philosophy

has given to relations has been inspired by Russell’s work. But Russell
was not the first philosopher to give relations significant philosophical
scrutiny. No other tradition has paid relations as much attention as the
medieval and early modern scholastic tradition, in which nearly every
theologian and philosopher devoted some ― and often considerable
― discussion to relations.2 One does not have to look far to see why.
Sydney Penner This is a tradition with a healthy respect for its intellectual forebears,
and two of the most prominent of those forebears were Aristotle
Merton College
(the Philosopher) and Augustine. Relations form one of the ten
University of Oxford
basic categories of being for Aristotle.3 A further six categories look
suspiciously like they might involve relations. It is hardly surprising,
then, that an Aristotelian tradition would take note of relations. If
1. “Logical Atomism”, in Logic and Knowledge: Essays, 1901–1950, ed. Robert
Charles Marsh (London: Allen and Unwin, 1956), 333.
2. Assertions such as the following by D.M. Armstrong appear rather ill-ad-
vised: “Philosophy has been a long time coming to grips with the category
©  2013  Sydney Penner of relations…. It is not until the late nineteenth and the twentieth century
with C.S. Peirce, William James, and Bertrand Russell that relations begin (no
This work is licensed under a Creative Commons more than begin) to come into focus” (Universals: An Opinionated Introduction
Attribution-NonCommercial-NoDerivatives 3.0 License. [Boulder, CO: Westview Press, 1989], 29).
<www.philosophersimprint.org/013002/> 3. Aristotle, Categories 7 6a37–8b24 and Metaphysics V.15 1020b26–1021b11.

 sydney penner Suárez on the Reduction of Categorical Relations
that weren’t motivation enough, Augustine made relations central
to discussions of the Trinity by suggesting that the divine persons
are constituted in some way by relations.4 This combination of
philosophical and theological motivation gave rise to a rich tradition
of sophisticated thought about relations.
This tradition remains relatively unexplored. Some valuable
work has been done on scholastics from the twelfth century to the
early fourteenth century.5 The next philosopher whose work receives
significant attention is Gottfried Wilhelm Leibniz (1646–1716), heir
to the medieval discussions.6 Both endpoints have further work left
to be done, but in between we have several centuries almost entirely
unaccounted for. Francisco Suárez (1548–1617), one of the shining
philosophical lights of Spain’s Golden Age, falls into that unexplored
period.7 He wrote more about relations than anyone else of whom I
know.8 The centrepiece of his discussion is a book-length treatise
4. De Trinitate V.
5. The central English-language works are by Jeffrey E. Brower, “Abelard’s Theory
of Relations: Reductionism and the Aristotelian Tradition”, The Review of Meta-
physics 51 (1998): 605–31; “Relations Without Polyadic Properties: Albert the
Great on the Nature and Ontological Status of Relations”, Archiv für Geschichte der
Philosophie 83 (2001): 225–57; “Medieval Theories of Relations”, The Stanford En-
cyclopedia of Philosophy (Winter 2010 Edition), ed. Edward N. Zalta, URL = <http://
plato.stanford.edu/archives/win2010/entries/relations-medieval/>; and by
Mark G. Henninger, Relations: Medieval Theories 1250–1325 (Oxford: Clarendon
Press, 1989). Some additional works will be cited in the course of the paper.
6. I will not cite the large literature on the question of whether Leibniz was a
reductionist about relations; interested readers can consult Anja Jauernig’s
“Disentangling Leibniz’s Views on Relations and Extrinsic Denominations”
(Journal of the History of Philosophy 48 [2010]: 171–205) for references to the
relevant literature, not to mention reading her paper for its own excellent
discussion of the question.
7. For an introduction to Suárez, see Jorge J.E. Gracia’s “Francisco Suárez: The
Man in History” (American Catholic Philosophical Quarterly 65 [1991]: 259–66)
and the other articles in the same journal issue. Alfred J. Freddoso provides
a nice overview of Suárez’s metaphysics, albeit with an eye to efficient causa-
tion, in his introduction to On Creation, Conservation, and Concurrence: Meta-
physical Disputations 20–22 (South Bend, IN: St. Augustine’s Press, 2002).
8. The qualification “of whom I know” is necessary, given the state of our knowl-
edge of the period.
philosophers’ imprint –  2  – vol. 13, no. 2 (january 2013)
entitled On Real Relation,9 the 47th of his 54 Metaphysical Disputations.
He also deals with relations extensively elsewhere ― e. g., with divine
relations in De Deo Uno et Trino and with relations of reason in the
54th of the Metaphysical Disputations ― but On Real Relation will provide
ample material for present purposes.10
The history of discussion about relations to which Suárez is heir is
both a blessing and a curse, especially to someone of his temperament.
Suárez is extraordinarily scholarly, painstakingly judicious with his
sources, and a harmonist who finds it well-nigh impossible to throw
received parts of the tradition overboard. On the one hand, it would be
a challenge to find any interesting argument about relations from the
previous two millennia of Western philosophical reflection that does
not receive at least passing attention from Suárez. On the other hand,
as we will see later in the paper, there are various places where one
gets the impression that Suárez ties himself into knots in increasingly
desperate attempts to preserve parts of the tradition that he might
better have jettisoned.11
9. About 55,000 words in Latin, to be more precise. While writing my paper I
was not aware of any other English-language discussions of Suárez on rela-
tions, but another paper has just been published: Jorge Secada, “Suárez on
the Ontology of Relations”, in Interpreting Suárez: Critical Essays, ed. Daniel
Schwartz (Cambridge: Cambridge University Press, 2011), 62–88. Secada
covers some of the same material that I do, though our emphases differ. On
reading his paper, I was gratified to see that our interpretations appear to be
broadly in agreement.
10. I will cite Metaphysical Disputations (henceforth: DM) by disputation, section,
and paragraph number. Latin quotations are from the 1597 Salamanca edition,
which provides the most reliable text. The work is most easily available, how-
ever, as volumes 25 and 26 of the Opera omnia published in Paris in 1856–78
and reprinted in Hildesheim in 1965. Quotations from other works are from
the Opera omnia and are cited by the relevant division numbers. All transla-
tions are mine, with consultation of John P. Doyle’s valuable translation of the
47th disputation (On Real Relation [Milwaukee: Marquette University Press,
2006]) in relevant cases.
11. Cf. Karl Eschweiler’s comment: “Das eigentümlich Neue dieser [jesuitischen]
Schulphilosophie … wird bei Suarez noch gern im Kompromiß mit den aus der Hoch-
scholastik überkommenen Denkgewohnheiten vorgetragen” (“Roderigo de Arriaga
S. J.: Ein Beitrag zur Geschichte der Barockscholastik”, Gesammelte Aufsätze zur
Kulturgeschichte Spaniens 3 [1931]: 255).
 sydney penner Suárez on the Reduction of Categorical Relations
The paper will proceed as follows: Section 1 presents some of the
relevant terminology and taxonomy for talking about Suárez’s account
of relations. In section 2, I introduce the motivation for a reductionist
account of relations and look at Suárez’s main treatment of that question
as well as a related question about whether relations constitute an
Aristotelian category. Sections 3 and 4 take a closer look at Suárez’s
reductionism, especially some of the reasons why he is reluctant to
throw in his lot with other reductionists wholeheartedly. Section 5
sketches some worries concerning reductionism about relations and
reveals that Suárez’s reductionism may not be as thoroughgoing as it
appears at first glance. I end by addressing a lingering issue concerning
relations’ status as a distinct category of real being.
1.  Terminological and taxonomic preliminaries
Medieval philosophers quickly developed a professional vocabulary
for talking about relations. Suárez shares this standard vocabulary
with others in his philosophical tradition, but it may sound unfamiliar
to modern ears, so a quick review is in order. I will point out a few of
Suárez’s philosophical positions along the way.
First, there are the various terms for relations: ‘relation’ (‘relatio’),
‘comparison’ (‘comparatio’), ‘respect’ or ‘outward-looking-ness’
(‘respectus’), ‘ordering’ (‘ordinatio’), and so on.12 Some of these terms
have connotations that make them more appealing to advocates
of certain positions than to others. ‘Comparison’, for example,
has psychological connotations that make it a preferred term for
conceptualists about relations. Suárez seldom uses ‘comparison’.
The terms that he uses repeatedly are ‘relation’, ‘respect’, and
‘habitude’ (‘habitudo’). A challenge with this abundance of terms
for relations is determining when a given author uses them as
synonymous expressions and when different terms signal different
concepts. I see no reason to worry about Suárez’s use of ‘relation’
and ‘respect’. The problematic term is ‘habitude’ (which is why I
12. See Brower, “Medieval Theories of Relations”, §2.1.
philosophers’ imprint –  3  – vol. 13, no. 2 (january 2013)
prefer to use the archaic English cognate rather than a more elegant
translation). In some texts it looks like Suárez may be using it as just
another term for relations. In other texts, however, one senses that
there may be a distinction lurking in the background, a distinction
that he, unfortunately, does not spell out. What is clear is that he
primarily uses ‘habitude’ when he has transcendental relations in
mind (see the taxonomy below). Since the focus in this paper is on
categorical relations, just what Suárez means by ‘habitude’ is less of
a concern for us.13
He also frequently uses the term ‘relative’ (‘relativum’), which in its
strict sense he defines as something that consists of a subject together
with its relation — so Peter, with his relation of similarity to John, is a
relative. Suárez notes that the term is also used more loosely to refer
to the subjects of relations.14
Second, there are the terms for the different elements needed to
construct a theory of relations. Let’s distinguish several terms with
reference to the example of Peter’s relation of similarity to John. Let’s
say they are similar in both being white.
The subject of the relation is Peter. The view is that relations are
accidents and accidents require subjects of which to be accidents.
There is a further question in the vicinity: namely, what special claim
does Peter have to being the subject of this relation? Why not John? Or,
better, both Peter and John?
13. Suárez’s predecessors disagree about how to use these terms. Henry of Ghent,
for example, distinguishes between ‘relation’ and ‘habitude’; see Jos Decorte,
“‘Modus’ or ‘Res’: Scotus’ Criticism of Henry of Ghent’s Conception of the Re-
ality of a Real Relation”, in Via Scoti: Methodologica ad mentem Joannis Duns
Scoti, ed. Leonardo Sileo (Rome: Edizioni Antonianum, 1995), vol. 1, 415–18,
and Scott Williams, “Henry of Ghent on Real Relations and the Trinity: The
Case for Numerical Sameness Without Identity”, forthcoming in Recherches de
Théologie et Philosophie Médiévales. Scotus, on the other hand, says in Ordinatio
II, dist. 1, q. 5 (p. 115 in vol. 7 of the Vatican edition) that ‘respect’, ‘relation’, and
‘habitude’ are synonymous.
14. DM XLVII.6.2.
 sydney penner Suárez on the Reduction of Categorical Relations
Most philosophers from Avicenna to Leibniz take it as axiomatic
that an accident cannot have more than one subject.15 Given that
premise, it follows quickly enough that relations ― being accidents
― cannot have more than one subject and thus cannot inhere in both
relata.16 The standard medieval view, then, is that, in ordinary cases,
one has a pair of converse relations, one of which inheres in a as its
subject and terminates in b and the other of which inheres in b and
terminates in a.17 This is true both for asymmetrical relations (e. g.,
Peter’s being taller than John and John’s being shorter than Peter) and
for symmetrical relations (e. g., Peter’s being similar in colour to John
and John’s being similar in colour to Peter).18 The claim that an accident
can have only one subject strikes me as worth questioning, but doing
15. For some helpful references, see Massimo Mugnai, Leibniz’ Theory of Relations
(Stuttgart: Franz Steiner, 1992), Chapter 2. There are a few dissenting voices.
In the thirteenth century, Nicholas of Paris accepts the one-subject rule in a
sense but thinks that, in the case of relations, multiple things together func-
tion as a single subject. On his view, the relation of paternity is identical to the
relation of filiation, and the relation is in the father and son together as in one
subject. In the seventeenth century, Rodrigo de Arriaga also holds that con-
verse relations are identical, so he has to accept either Nicholas of Paris’s view
or that relations have more than one subject. For more on Nicholas of Paris,
see Heine Hansen, “Strange Finds, or Nicholas of Paris on Relations”, in Logic
and Language in the Middle Ages: A Volume in Honour of Sten Ebbesen, ed. Jakob
Leth Fink, Heine Hansen, and Ana María Mora-Márquez (Leiden: Brill, 2012).
On Arriaga, see my “Rodrigo de Arriaga on Relations”, The Modern Schoolman
89 (2012): 25–46.
16. See Suárez’s reply to Henry of Ghent in DM XLVII.6. It is not clear that the
view that Suárez attributes to Henry is actually Henry’s view, but the reply
certainly elucidates Suárez’s position.
17. For more on this as a standard part of the medieval picture, see Henninger,
Relations, 4–6, and Brower, “Medieval Theories of Relations”, §3.1. We might,
following Bertrand Russell, call the standard medieval view ‘monadistic’,
though Russell defines monadism in terms of relational propositions and ad-
jectives rather than in terms of relations and accidents. See his Principles of
Mathematics (London: Routledge, 2010), Chapter 26.
18. Suárez’s term for symmetrical relations is ‘relations of equiparence’ (‘relatio-
nes aequiparantiae’) and for asymmetrical relations is ‘relations of disequipa-
rence’ (‘relationes disquiparantiae’). See DM XLVII.15.2.
philosophers’ imprint –  4  – vol. 13, no. 2 (january 2013)
so now would take us too far afield.19 For the moment, we should just
accept, as a standard part of the medieval picture, the assumption that
each relation has a single subject.
The foundation of the relation is Peter’s whiteness. This is the feature
in John that makes it the case that he is similar to Peter. The thought
here is straightforward enough. Whenever we have a subject and its
relation, there is something in virtue of which that relation holds of
that subject. That something is the foundation for the relation.20 The
foundation of a relation is also that from which the relation derives its
reality, but this is to get ahead of the story.
Matters become more complicated when we introduce the term
‘reason of founding’ (‘ratio fundandi’), a term that seems to have become
a standard part of the later scholastic discussions. It is not at all clear,
however, what the term contributes to the discussion — i. e., what
the difference is supposed to be between a reason of founding and
a foundation. In fact, one of the views defended by some scholastic
philosophers is that there is no difference. Others, however, argue
that there is a difference. Suárez himself ― in a rather unsatisfactory
discussion of the matter ― concludes that there is a reason of founding
in every foundation but that the reason of founding is not “really
or physically distinct” from the foundation. Rather, they are only
“distinguished metaphysically or in reason in order to explain the
rationes of things.”21 One gets the impression that Suárez is not sure of
himself in this passage.
He is more sure of himself with respect to the question of whether the
subject and foundation of a relation need always be distinct. Aquinas
seems to have thought that they must always be distinct.22 One might
think that the foundation for a relation must itself be an accident, since
19. I question it in my “‘An Accident that is Simultaneously in Two Subjects’:
Leibniz and Some Predecessors on the Possibility of Two-Subject Accidents”
(unpublished).
20. DM XLVII.7.
21. DM XLVII.7.14.
22. Cf., e. g., Summa contra Gentiles IV.14.12.
 sydney penner Suárez on the Reduction of Categorical Relations

a relation is an accident and is supposed to derive its reality from its these terms. We made a distinction on the subject side — i. e., between
foundation. Suppose we have a relation inhering in a subject that is a subject and foundation — but made no parallel distinction on the
substance where we are tempted to say that the subject and foundation terminus side. Despite not typically making this distinction in his
are identical in reality. If the subject and foundation are identical, the discussions, Suárez is not opposed to making it. As he himself notes:
foundation must also be a substance. Since the being of the relation
For, just as a white thing is apt for founding a relation of
is derived from that of its foundation, it must also have the being of
similarity because it has whiteness and that whiteness itself
a substance. But relations have the being of accidents. Therefore, it
is an apt foundation because it has such a formal unity, so
must not be the case that the subject and foundation are identical.23
also the same white thing is apt for terminating the relation
Suárez, however, rejects this argument, noting that there are cases
of another [white] thing because it has a whiteness of the
where we have good reason to say that the relation arises from the
same character or unity with another whiteness.26
essence of the subjects and not from some accident that they might
have. For example, John and Peter also share the relation of specific
Of course, just as in the case of the foundation, the terminus will not
identity — i. e., Peter is the same species as John. But it is implausible to
always be distinct in reality from that which provides the aptness for
say that they are of the same species as a result of some accident — e. g.,
terminating. That is, in some cases the essence of the terminus will
whiteness. After all, any number of things are similar in that respect
itself provide the aptitude for termination.
without being of the same species as a consequence. Rather, it seems
On to matters taxonomic. Here is a diagram of how Suárez charts
clear that John and Peter are of the same species because they have the
the territory:
same essence.24 In response to the argument on behalf of Aquinas’s
view, Suárez denies the premise that says that if the being of a relation
is derived from a substance, it must consequently itself have the being
of a substance. He notes that there are other cases as well where
accidents — e. g., duration — are distinguished from substances only by
reasoned reason (ratione ratiocinata).25 So Suárez parts company with
Aquinas on this question and concludes that subjects and foundations
are not always really distinct.
The terminus of the relation in question is John. That termini are
needed is obvious, since it is of the essence of a relation to be toward
another thing. That other thing is the terminus. This diagram is somewhat misleading in that Suárez does not think
Those, then, are the ubiquitous terms in scholastic discussions of that ‘relation’ is univocal between real relations and conceptual
relations. An attentive reader, of course, will notice an asymmetry in
26. DM XLVII.16.18: “…nam sicut album aptum est ad fundandam relationem si-
23. Suárez suggests this argument on behalf of his opponents in DM XLVII.7.2. militudinis, quia habet albedinem, et ipsa albedo est aptum fundamentum,
24. DM XLVII.7.5. quia habet talem unitatem formalem, ita etiam idem album est aptum ad ter-
minandam relationem alterius, quia habet albedinem ejusdem rationis vel
25. DM XLVII.7.8. unitatis cum albedine alterius.”

philosophers’ imprint –  5  – vol. 13, no. 2 (january 2013)

 sydney penner Suárez on the Reduction of Categorical Relations
relations,27 but I think it is clear what is meant by this classification. It
is also the way that Suárez himself presents it.28
Categorical relations are the relations that fall into the category of
relation, where the category is one of the highest genera of being.
Paradigmatic examples of relations (similarity, equality, knowledge,
spatial relations) are supposed to fall into this category. This kind of
relation is the subject of the remainder of the paper, so I will not say
more about them here by way of introduction.
Transcendental relations are somewhat puzzling both in Suárez and
in scholastic thought more generally. Scholastic authors not only give
very different accounts of transcendental relations, disagreeing about
basic points such as whether they are real or not — they disagree
about what feature determines the extension of the class. Suárez
offers two characterizations. In one place, he says that transcendental
relations run through all the categories (that is, they transcend the
categories);29 in other places, he says that transcendental relations are
those relations that are essential to something — i. e., relations that
are intrinsically included in the concepts of things.30 It is not clear,
however, that these two characterizations capture the same class. His
contemporary Antonio Rubio (1548–1615), for example, thinks that
not all relations that are essential transcend the categories. Since he
gives priority to the feature of transcending the categories, he thinks
that only some transcendental relations are essential.31 Rubio takes
a minority view, however. The core notion is usually that of being
essential to the subject. A transcendental relation is one without
which a thing could not be the thing that it is.32 Despite his offering
27. DM XLVII.3.3.
28. DM XLVII.3.
29. DM XLVII.3.10.
30. E. g., DM XLVII.1.5 and 3.12–13.
31. Logica mexicana, hoc est commentarii breviores et maxime percipue in universam Ar-
istotelis dialecticam (Lyons: Antonio Pillehotte, 1620), 358.
32. Transcendental relations thus are akin to one of the key notions of internal
relations in discussions of monism and idealism. See, for example, Jonathan
philosophers’ imprint –  6  – vol. 13, no. 2 (january 2013)
of two characterizations, I think it is the one about essentiality that
drives Suárez’s discussion.33
Regardless the disagreement about how to characterize
transcendental relations, there does seem to be some agreement
about what some of the paradigmatic examples of transcendental
relations are. One is the essential dependence that every created thing
has with respect to God. Both of Suárez’s characterizations apply to
this example. Beings in all the categories depend on God in this way,
and that dependence is taken to be essential to everything in those
categories. Another widely used example, and one that Suárez favours,
is the relation of inherence that an accident has to its subject. This
relation seems to transcend most but not all of the categories; it is,
however, plausibly essential to its subjects.
There are further issues that are unclear about Suárez’s account of
transcendental relations. To be fair, transcendental relations are not
the focus of his discussion; despite the title of the treatise being On
Real Relation, his focus is on one kind of real relation — namely, the
categorical kind. Categorical relations, of course, were the usual focus
in medieval discussions of relations and are supposed to encompass
the paradigmatic cases of relations. Furthermore, for at least some
philosophers, transcendental relations are not real relations and hence
are justifiably sidelined in discussions about relations. Suárez, however,
does think that transcendental relations are real. Not only that — it
quickly becomes clear that they play crucial roles in central parts
of his philosophy. Causation, for example, involves transcendental
relations.34 Hence, a clear explication of transcendental relations
Schaffer, “The Internal Relatedness of All Things”, Mind 119 (2010): 349–50.
33. I do not know of any fully satisfactory discussion of transcendental relation
with respect to Suárez or otherwise, but readers may wish to consult G.M.
Manser, “Begriff und Bedeutung der transzendentalen Beziehung”, Divus
Thomas 4 (1941): 351–60. Although also not proffering a full investigation,
Secada says more than I do about Suárez’s account of transcendental rela-
tions in “Suárez on the Ontology of Relations”, 77–80.
34. See DM XLVIII.1 and J. Patout Burns, “Action in Suárez”, The New Scholasticism
38 (1964): 453–72.
 sydney penner Suárez on the Reduction of Categorical Relations
would be desirable. At present, however, we just need to keep in mind
that Suárez’s account of categorical relations is not the full story. There
are other real relations lurking in the vicinity, a point that becomes
especially relevant when we see that it is not obvious that Suárez’s
reductionism extends to these other relations.
Conceptual relations constitute one of the three main divisions of
conceptual beings or beings of reason (the other two being negations
and privations).35 Conceptual relations are analogous to real relations
but fail to be real, because one of the conditions for real relations is
not met. For example, Suárez, following the standard medieval view,
thinks that self-identity is not a real relation, because there is no real
distinction between the relata. Relations between relata that are not
real also fail to be real relations and so are conceptual relations.
2.  Of minimal entity
Suárez opens On Real Relation by noting that Averroes says that, among
all the categories of being, relations have minimal entity.36 The general
sentiment goes back to Aristotle, who says that “the relative is least
of all things a real thing or substance”.37 Now, talk of something that
has a little bit of being but not as much being as most other things
can quickly become puzzling, given that the claim is not simply that
relations have less mass than other things or something like that. But
perhaps we can see why one might be tempted to say that relations
have minimal entity.
Start with the thought that we often express propositions with
relational predicates: ‘Peter is five metres away from John.’ If the
35. Suárez devotes the 54th and last Metaphysical Disputation to beings of rea-
son. Conceptual relations are covered in section 6 of this article. On Suárez
on beings of reason, see Christopher Shields, “Shadows of Beings: Francisco
Suárez’s Entia Rationis”, in The Philosophy of Francisco Suárez, ed. Benjamin Hill
and Henrik Lagerlund (Oxford: Oxford University Press, 2012), 57–74.
36. DM XLVII.proem.
37. Metaphysics N.1 1088a23–24. Quotations from Aristotle’s Metaphysics are from
W.D. Ross’s translation in The Complete Works of Aristotle: The Revised Oxford
Translation, ed. Jonathan Barnes (Princeton: Princeton University Press, 1984).
philosophers’ imprint –  7  – vol. 13, no. 2 (january 2013)
proposition is true, there should be some being or beings in the
world that make it true. So we might be tempted to think that there
really is a relation or relations of distance in the world that makes
the proposition ‘Peter is five metres away from John’ true. After all,
Peter by himself does not make the proposition true, nor does John.
So maybe a relation is called for. On the other hand, if we had a world
with nothing but Peter and John in it, five metres apart, it is not too
hard to imagine that someone counting the things in the world would
say that it had two things in it.38 In fact, it is not too difficult to find
sophisticated philosophers who have forgotten about relations when
doing their ontological enumerations. Even after we have granted that
there must be a relation around in order to make our aforementioned
proposition true, we may still think this relation a rather ghostly being.
Just where is it? How can I get my hands on it?
Such thoughts may have led to the slogan that relations have
minimal entity. Such thoughts no doubt also motivate reductionist
programmes. One of the most heavily debated questions in medieval
discussions about relations was whether relations can be reduced to
absolute things (absolute accidents such as qualities and quantities,
usually). If Peter and John and their absolute accidents are enough to
make relational propositions about them true, then it is no wonder that
our attempts to find some additional relational thing left us feeling like
we were chasing a ghost. Choice of example matters here. It is more
difficult to see how this reduction should go in the case just mentioned.
But the stock medieval example of a relation ― the relation of similarity
between two white things ― lends itself to such reductionism. It seems
plausible that Peter’s whiteness and John’s whiteness are enough to
make the proposition ‘Peter is similar in colour to John‘ true. Thinking
of the latter sort of example, then, might lead us to try formulating a
general reductionist programme for relations.
The question of reduction is at the forefront of Suárez’s mind right
from the start of the treatise. The questions for the first two sections
38. We can ignore for the present the possibility that we should also count a good
many qualities and the like.
 sydney penner Suárez on the Reduction of Categorical Relations
are (1) “whether relation is a true genus of real being distinct from
the rest of the categories” and (2) “whether real categorical relations
are really or modally ― that is, ex natura rei ― distinct from substance
and all absolute accidents”.39 The second question is of more obvious
interest, since it seems to get to the heart of the reductionist question,
while the first one more concerns fitting relations into an Aristotelian
categorical scheme. But the two questions intertwine and so need to
be examined together. Sections in Suárez’s works are usually more or
less stand-alone pieces. It is unusual, then, that the first section ends
on a very unsatisfying note, with a number of its opening arguments
receiving their replies only in the second section.40 Let’s look at the
two sections in turn.
2.1 A distinct genus of real being?
An affirmative answer requires that two conditions be met: (1) that
relations have real being and (2) that relations constitute a distinct
category. So three positions present themselves.
The first position denies that the first condition is met. Advocates
of this position argue that relations do not form a distinct genus of real
being, because relations are not real beings. Only absolute things exist
in extramental reality. Suárez attributes this view to Peter Aureoli (ca.
1280–1322).41
Suárez presents three arguments for doubting that the first
condition is met. The first one is rather complex and not especially
relevant for our purposes.42 But the latter two are straightforward, and
Suárez’s response to them will be illuminating with respect to his
own position. One we might call the Cambridge Change Argument. This
argument goes back at least to Aristotle43 and has been popular ever
39. DM XLVII.1–2.
40. The arguments from §1.1–4 are addressed in §2.24. The arguments from §1.5–
7 are left until even later: §4.17–20.
41. DM XLVII.1.8. For Aureoli on relations, see Henninger, Relations, 150–73.
42. It can be found in DM XLVII.1.1.
43. Physics V.2 225b11–13.
philosophers’ imprint –  8  – vol. 13, no. 2 (january 2013)
since.44 It starts from the observation that relational predicates can go
from applying to not applying to a subject without any real change in
the subject. Suppose a is similar in colour to b (they are both white). If
we now paint b red, a will no longer bear the relation of similarity to b.
Yet it seems that only b changed, not a. Hence, the thought goes, the
relation of similarity added no real being to a, for if it had, then either
a would really change merely because something else got painted or
a would still bear the relation of similarity even though b is no longer
the same colour. Neither option is attractive, so relations must not add
any real being to subjects.45
The other we might call the Necessary Consequence Argument. Here
is how Suárez puts it:
For if we posit, for example, two white things in reality,
they, by force of the absolute qualities that they have, are
similar to each other. Therefore, they have that relative
denomination by force of the absolute things existing
together or being taken together, without any other
real addition. Therefore, a relation adds nothing real to
absolute things.46
The relation necessarily follows with the positing of the two white
things, so why posit some additional entity to explain the relation?
Better just to take the relation as an “ontological free lunch”, as D.M.
Armstrong puts it.47 That, of course, is not quite enough to say that
relations are not real. Relations might fail to add any real being to
absolute things because they are reducible to absolute things rather
44. See Henninger, Relations, 8–9, for some references.
45. DM XLVII.1.2.
46. DM XLVII.1.3: “…quia si ponamus v. g. duo alba in rerum natura, illa, ex vi qualitatum
absolutarum quas habent, sunt inter se similia: ergo habent illam denominationem rela-
tivam ex vi absolutorum simul existentium, seu simul sumptorum absque aliqua alia
additione reali: ergo relatio nihil rei addit rebus absolutis.”
47. Armstrong thinks this is true of only some relations — namely, those he calls
internal relations (Universals, 100).
 sydney penner Suárez on the Reduction of Categorical Relations
than because they do not really exist. That is, the Necessary Consequence
Argument could just as well lead to reductionism as to conceptualism.
That possibility is not taken into consideration here, but noting it
shows us why this first question about whether relations form a
distinct category of real being cannot be answered fully before tackling
the reducibility issue in the second section of On Real Relation.
So those are two ways of arguing that the first condition ―
namely, that relations have real being ― is not met. Suárez is clearly
sympathetic to these arguments, but we will see how he responds to
them later.
The second position denies that the second condition is met.
Suárez reports that Domingo de Soto attributes this position to Zeno
of Elea and other Presocratics.48 Advocates of this position concede
that relations are real but deny that they constitute a distinct category
of being. The thought, roughly, is that relations are a transcendental
condition of being ― just as being in a genus is a condition or
property of all beings, so being related is a condition or property of
all beings49 ― and so can be found in all beings rather than making
up a distinct category. In Suárez’s taxonomy, this amounts to the
claim that all real relations are transcendental relations: there are
no additional relations to be put in a special category. Obviously,
this issue will quickly land us in the tangle about transcendental
relations; fortunately, sorting this out is not crucial for present
purposes. Suffice it to say that Suárez rejects this position and claims
that there are categorical relations and that they are distinct from
transcendental relations.50
Finally, the third position claims that relations meet both
conditions and so relations do form a distinct genus of real being.
48. DM XLVII.1.9.
49. For this analogy, see Domingo de Soto, In Porphyrii Isagogen, Aristotelis Catego-
rias, librosque De demonstratione, absolutissima commentaria (Venice: apud Guer-
raeos fratres et socios, 1573), 195.
50. The arguments against the second condition being met are in DM XLVII.1.5–7;
the main part of Suárez’s response comes in 4.17–20.
philosophers’ imprint –  9  – vol. 13, no. 2 (january 2013)
This is the traditional view, and Suárez marshals an impressive list
of philosophical authorities on its behalf.51 He also raises several
traditional arguments for the view but is unsatisfied with each.
One argument comes from theology: the three divine persons are
constituted by three relations, and these relations clearly are not
just mental entities. But if we admit that the relations in the case of
God are real, it seems probable that the relations in the case of the
created order are also real. Suárez, however, thinks the disanalogy
between the cases renders the argument ineffective.52 A second
argument argues from the linguistic data. Relative predications are
common, and we do not merely imagine that things are related.
Hence, there must be some extramental foundation for them. But
absolute things cannot ground relative predications, so they must
be grounded in relative things. But, as Suárez points out, this simple
argument does not settle the matter. After all, it begs the question
against reductionism about relations, for the idea that absolute
things can ground relative predications is just what the reductionist
is committed to.53 Similar concerns weigh against a third argument
from the order of the universe. The reductionist agrees that there is
mind-independent order in the universe, but she denies that we need
relational things for that order in addition to absolute things.54
Suárez ends the first section, however, by affirming the third
position. But he says that so far he primarily has the weight of authority
on his side. He seems to think that the argument from the linguistic
data carries some force, but he grants that the reductionism question
needs to be tackled first before it can be evaluated properly.55 That is
the question we turn to next.
51. DM XLVII.1.10.
52. DM XLVII.1.11 and 13.
53. DM XLVII.1.12–13.
54. DM XLVII.1.14.
55. DM XLVII.1.13 and 15.
 sydney penner Suárez on the Reduction of Categorical Relations
2.2 Distinct from absolute things?
We are used to talking about whether one class of things can be reduced
to another class of things — e. g., whether mental phenomena can be
reduced to physical phenomena. Medieval philosophers get at the
same issue by asking what sort of distinction holds between entities of
the two classes. So, in the second question of On Real Relation, Suárez
asks whether categorical relations are really or modally distinct from
absolute things.
A brief digression into Suárez’s theory of distinctions is in order
here. What kinds of distinctions one can have in one’s conceptual
toolbox was itself a matter of heated dispute, and Suárez ― is it any
surprise? ― has a treatise on the subject. Suárez recognizes three main
kinds of distinction, with two variations of one of them.
• At one end is the real distinction. In line with traditional
views, Suárez takes separability to be the sign of a real
distinction. In contrast to some — e. g., Scotus — Suárez
requires mutual separability for a real distinction. That is,
the sign of a real distinction between a and b is that a can
exist without b and b can exist without a.56
• One-way separability is the sign for an intermediate kind
of distinction, the modal distinction. If either a or b can exist
without the other but not the other way around, we can
infer a modal distinction between a and b.57
• If a and b are neither mutually nor one-way separable, then
all that we can have is a distinction of reason.58 Distinctions
of reason, however, come in two flavours, according
to Suárez. The distinction of reasoning reason has no
foundation in reality, but the distinction of reasoned reason
does, even though there is no real or modal distinction in
56. DM VII.2.9–21.
57. DM VII.2.6–8.
58. DM VII.2.28.
philosophers’ imprint –  10  – vol. 13, no. 2 (january 2013)
reality to provide that foundation.59 The classic example
of a distinction of reasoned reason is between the justice
and the mercy of God. According to the doctrine of
divine simplicity, there are no real or modal distinctions
within God; nevertheless, there is supposed to be some
extramental foundation for our attributing justice and
mercy as separate attributes to God.
Suárez does not think that any further kinds of distinction beyond
the basic three kinds are possible. The Scotist formal distinction,
for example, is acceptable only if it turns out to be the same as the
distinction of reasoned reason or the modal distinction.60 These may
exhaust the kinds of distinctions that Suárez grants; they do not
exhaust the vocabulary he uses. In particular, he sometimes uses the
terms ‘actual distinction’ and ‘distinction ex natura rei’. I take both of
those terms as covering both real and modal distinctions but excluding
merely mental distinctions. It is, however, not always entirely clear
which distinction Suárez has in mind, despite admirably spelling out
his terminology in a treatise devoted to distinctions.
Given Suárez’s account of distinctions, we can see that a negative
answer to the second section’s question amounts to reductionism.
That is, if there is no real or modal distinction between relations
and absolute things, then all that is left is a distinction of reason.
This distinction of reason might have some sort of foundation in
59. DM VII.1.4–5.
60. DM VII.1.13–21. For more on Suárez on distinctions, see Richard Glauser,
“Descartes, Suarez, and the Theory of Distinctions”, in The Philosophy of Marjo-
rie Grene, ed. Randall E. Auxier and Lewis Edwin Hahn (Chicago: Open Court,
2002), 417–45; Stephen Menn, “Suárez, Nominalism and Modes”, in Hispanic
Philosophy in the Age of Discovery, ed. Kevin White (Washington, D.C.: Catholic
University of America Press, 1997), 226–56; and Roger Ariew, “Descartes and
Leibniz as Readers of Suárez: Theory of Distinctions and Principle of Indi-
viduation”, in The Philosophy of Francisco Suárez, ed. Benjamin Hill and Henrik
Lagerlund (Oxford: Oxford University Press, 2012), 38–53. There is also a
dissertation on the subject but that I have not had a chance to consult yet:
Michael V. Murray, “The Theory of Distinctions in the Metaphysics of Francis
Suarez” (Ph.D. dissertation, Fordham University, 1945).
 sydney penner Suárez on the Reduction of Categorical Relations

reality — i. e., it might be a distinction of reasoned reason, but there
is no extramental distinction. Suárez does in fact deny that there is a
real or modal distinction between relations and absolute things and
thereby commits himself to a reductionist programme. (More precisely,
Suárez’s position, as we will see, may be that categorical relations can
be reduced to a base that includes no categorical relations but does
include more than just purely absolute things, but we can ignore that
complication for the moment.)
We might expect to see three positions on the table ― that there is
a real distinction, that there is a modal distinction, and that there is no
extramental distinction ― but Suárez, ever the careful scholar, goes
through six positions along with the arguments for each of them.61 We
can gloss over some of those subtleties and divide them broadly into a
reductionist position and the rest — i. e., the non-reductionist positions.
Whether a non-reductionist might take the distinction to be real or
modal and whether this is true for all relations or only certain classes
of them is not important for the moment.
The key weapon in any non-reductionist’s arsenal is the
Separability Argument in its various guises. The argument is so
standard in medieval discussions that Suárez usually just refers to
the “separability of a relation from its foundation” without bothering
to explain just how the argument is supposed to go. Here is one
way to formulate the argument: If a and b are identical, then it is
impossible for either to exist without the other. Foundations can exist
without their relations. Therefore, relations and their foundations
are not identical, or, more precisely, relations and their foundations
must be at least modally distinct. For example, the foundation of
Peter’s relation of similarity to John is Peter’s whiteness. But if John
were to get a dark tan, Peter would lose his relation of similarity but
might well still be white. Hence, Peter’s whiteness and his relation of
similarity cannot be identical.
61. They are presented in DM XLVII.2.2–21.
For part of his career, Suárez seems to have been moved by non-
reductionist arguments such as this. In a work published seven years
before the Metaphysical Disputations, he says:
But, indeed, I think there is some distinction ex natura rei
[between a relation and its foundation]: not, to be sure,
as of things that are wholly distinct, but as of different
formal rationes or as of a mode and the thing of which it
is a mode.62
He does not accept Scotus’s line that relations are things ― res ― really
distinct from absolute things, but he does here seem to accept that
there is a modal distinction between relations and their foundations.
Even in On Real Relation it is clear that the modal-distinction view still
has some pull on him. He explicates it at greater length than the other
views, offers more arguments in its favour, and says that this view is
“more probable” than most of the other views.63
But, in the end, he rejects the modal-distinction view. As Suárez
points out, the Separability Argument can be turned on its head:
We can turn back that argument about separability. For it
is impossible even with absolute power to make it so that,
for example, the two posited white things are not similar.
Therefore, this denomination [of similarity] is not taken
62. De Incarnatione Verbi (aka: Commentariorum ac disputationum in tertiam partem
Divi Thomae, tomus primus) q. 2, art. 8, n. 16 (= Opera omnia 17:328): “Ego vero
sentio esse quidem distinctionem aliquam ex natura rei, non quidem tanquam rerum
omnino distinctarum, sed tanquam formalium rationum diversarum, seu tanquam
modi, et rei cujus est modus.” Later in the same work (disp. 11, sect. 2, n. 16 [=
Opera omnia 17:440]), he says: “I respond that a relation is not in or united to
the subject or foundation as a distinct entity but as a mode …” (“Respondetur,
ipsam relationem non inesse vel uniri subjecto, seu fundamento, tanquam entitatem
distinctam, sed tanquam modum …”).
63. DM XLVII.2.7–11.

philosophers’ imprint –  11  – vol. 13, no. 2 (january 2013)

 sydney penner Suárez on the Reduction of Categorical Relations
from some thing or mode that is distinct ex natura rei from
both white things taken at the same time.64
The problem with the Separability Argument is that it only puts the
foundation in the subject of the relation in the putative reductive base
and then, of course, it turns out that foundations are separable from
their relations. But if we also include the termini, separability is no
longer so obvious (a point that Ockham emphasizes in his treatment
of relations). In fact, the Necessary Consequence Argument that we
looked at earlier precisely denies separability once the termini are in
the picture.
So Suárez throws in his lot with the reductionists. But here matters
become complicated. When Suárez first introduces the reductionist
view ― as the fifth on his list of six views and before he affirms it ―
he presents it simply as the view that denies that there is an actual
distinction between relations and absolute things; rather, this view
holds, there is only a distinction of reason with some foundation
in reality — i. e., a distinction of reasoned reason. Furthermore, he
immediately notes that this is the view that the nominalist school
teaches; he cites Ockham and Gregory of Rimini in particular.65 He
then goes on to credit Hervaeus Natalis with an especially good
presentation of the view. He even cites Aquinas for support, though
the passage he quotes is actually from a spurious work that may
also be by Hervaeus.66 But he clearly recognizes that this is the view
associated with the nominalists.
It is not difficult to see why. Consider the classic text for Suárez’s
presentation of his own view:
64. DM XLVII.2.15: “[P]ossumus retorquere argumentum illud de separabilitate, nam
positis, v. g. duobus albis, impossibile est etiam de potentia absoluta facere ut non sint
similia ergo haec denominatio non sumitur ex aliqua re aut modo reali distincto ex
natura rei ab utroque, albo, simul sumpto.”
65. DM XLVII.2.12.
66. DM XLVII.2.13. On the spurious work, see footnote 57 in John P. Doyle’s trans-
lation of the treatise.
philosophers’ imprint –  12  – vol. 13, no. 2 (january 2013)
But [relation] is not some thing or mode distinct ex
natura rei from every absolute form, but it is in reality
some absolute form, though not taken absolutely, but
taken as respecting another form, which the relative
denomination includes or connotes. Thus, for example,
similarity is some real form existing in the thing that
is denominated similar. It is not, however, distinct in
reality from whiteness, insofar as that which it posits
in the thing that is called similar is concerned, but only
insofar as the terminus that it connotes is concerned. And
thus similarity in reality is nothing other than that very
whiteness as respecting another whiteness as of the same
or similar ratio.67
Anyone familiar with medieval philosophy should hear, in this passage,
strains of Ockham’s theory of connotation as a tool for a programme of
ontological reduction. Ockham applies his theory of connotation much
more broadly than just to relational expressions, but he does apply it
to relational terms.68 As the story goes, we need to distinguish between
67. DM XLVII.2.22: “…illa vero non esse rem aliquam, aut modum, ex natura rei distinc-
tum ab omni forma absoluta, sed esse in re formam aliquam absolutam, non tamen
absolute sumptam, sed ut respicientem aliam, quam denominatio relativa includit seu
connotat. Ita ut similitudo, verbi gratia, aliqua forma realis sit existens in re quae de-
nominatur similis: illa tamen non sit in re distincta ab albedine, quantum ad id quod
ponit in re quae dicitur similis, sed solum quantum ad terminum quem connotat: et
ita similitudo in re non est aliud, quam ipsamet albedo ut respiciens aliam albedinem,
tanquam eiusdem seu similis rationis.”
68. In Summa logicae I.4, for example, he says that “all relative names are conno-
tative”. This text is available in translation as Ockham’s Theory of Terms: Part I
of the Summa Logicae, transl. by Michael J. Loux (Notre Dame, IN: University
of Notre Dame Press, 1974). For more on Ockham’s reductionist programme
and related issues, see Marilyn McCord Adams, William Ockham (Notre Dame,
IN: University of Notre Dame Press, 1987), 2 vols; Claude Panaccio, “Conno-
tative Terms in Ockham’s Mental Language”, Cahiers d’épistémologie no. 9016
(Montreal: Université du Québec à Montréal, Group de Recherche en Épis-
témologie Comparée, 1990); Paul Vincent Spade, “Ockham’s Distinctions Be-
tween Absolute and Connotative Terms”, Vivarium 13 (1975): 55–76, and “Ock-
ham, Adams and Connotation: A Critical Notice of Marilyn Adams, William
Ockham”, The Philosophical Review 99 (1990): 593–612; and Martin Tweedale,
 sydney penner Suárez on the Reduction of Categorical Relations
primary signification and secondary signification (or connotation).
Absolute terms signify only primarily. Connotative terms, on the other
hand, have both a primary and a secondary signification. In the case
at hand, the point is that we do not need to posit some new entity in
our ontology to be the primary significate of ‘similarity’. Instead, we
should just see that it is a connotative term that has the same primary
significate as ‘whiteness’ but differs from the latter term in also having
a secondary significate — namely, the whiteness of the terminus.
If we attend closely to Suárez’s text, however, it becomes clear
that he is not comfortable being bedfellows with Ockham and the
nominalists. He recognizes that the nominalist position is very close
to his own, but he warns readers repeatedly that his own view has
to be understood in a certain way in order to ward off unwanted
consequences. The unwanted consequences look like nominalist
consequences. The first indication of Suárez’s wariness comes after he
has presented the six different views and says which view he endorses:
Of these views, therefore, the one that seems most proven
to me is the fifth one, which Hervaeus and some other
Thomists taught, from the sense of which the nominalists
hardly differ at all …69
To say that the nominalists “hardly differ” from the affirmed view may
not seem significant, but recall that when the view was first presented,
the nominalists were the first exemplars of the view. Now there seems
to be something about them that makes Suárez hesitant. We need to
figure out what is worrying him.70
“Ockham’s Supposed Elimination of Connotative Terms and His Ontological
Parsimony”, Dialogue 31 (1992): 431–44.
69. DM XLVII.2.22: “Inter has ergo sententias mihi maxime probatur quinta, quam Her-
vaeus, et nonnulli alii Thomistae docuere, a quorum sensu fere nihil discrepant Nomi-
nales, et Aristoteles, et D. Thomas multum fauent vt vidimus.”
70. I should note that I am not offering a proper interpretation of Ockham’s ac-
count or that of any other nominalist. Whether any nominalist actually held
precisely the view that Suárez is worried about is not part of this paper’s am-
bit. Much of the nominalist tradition is, of course, severely underexplored, but
philosophers’ imprint –  13  – vol. 13, no. 2 (january 2013)
3.  Two desiderata
Suárez closes the second section of On Real Relation with two warnings:
(1) foundation and terminus do not contribute equally to a relation and
(2) sense must be made of the traditional distinction between mutual
relations and non-mutual relations.
3.1 Relations as one-subject accidents
Medieval reductionists commonly use the term ‘coexistence’. The
relation of similarity between Peter and John might be said to involve
the coexistence of two white things. The thought is a natural one.
As Scotus’s Separability Argument shows, the similarity relation is
separable from Peter’s whiteness; as his opponents point out, however,
it is not separable from two coexisting white things. This might suggest
that relations function in a way akin to collections. Just as each of the
goldfinches flying past my window contributes in the same way to the
flock, so both Peter and John (or their whitenesses) might be thought
to contribute in the same way to the relation of similarity. But Suárez
thinks this would be to get the analysis of relations wrong:
When it is said that a respective denomination arises
from the coexistence of multiple absolutes without
any real addition, that denomination should not be
understood as being taken equally and simultaneously
from multiple absolute forms, one intrinsic and the other
extrinsic [to the subject]. Rather, it should be understood
that this denomination indeed requires an association or
coexistence of such things or forms, but in each extreme
it is taken from that extreme’s own form as respecting
another form, which as such has the ratio of a relation,
readers interested in Ockham’s account of relations can consult Henninger,
Relations, Chapter 7; Marilyn McCord Adams, “Relations, Inherence, and Sub-
sistence: or, Was Ockham a Nestorian in Christology?”, Noûs 16 (1982): 62–75;
and Beatrice Beretta, Ad aliquid: la relation chez Guillaume d’Occam (Fribourg:
Éditions Universitaires Fribourg Suisse, 1999).
 sydney penner Suárez on the Reduction of Categorical Relations
even though in reality it is nothing other than the absolute
form itself.71
It is true that Peter’s relation of similarity to John requires the
coexistence of two forms — namely, Peter’s whiteness and John’s
whiteness. But the two forms do not contribute equally to that relation.
In Suárez’s view there are two relations here, each an accident and
each with one subject: Peter’s similarity to John and John’s similarity
to Peter. In order to get two relations instead of just one in the way we
get one flock of goldfinches, we need to identify relations with their
foundations in the subjects.
The point is most easily seen in the case of asymmetrical relations.
Suppose Peter is 200 centimetres tall and John is 180 centimetres tall.
So Peter is taller than John, the converse relation of which is not that
John is taller than Peter but that John is shorter than Peter. Suárez does
not think that the way to analyze either of these converse relations
is as the mere coexistence of Peter’s height and John’s height. That
would not capture the difference between the two converse relations.
Rather, Peter’s being taller should be identified with his height — not,
to be sure, with his height considered in itself, but with his height as
respecting John’s height. John’s being shorter, in turn, is to be identified
with John’s height as respecting Peter’s height.
The same analysis can, of course, then easily be made in the case of
symmetrical relations. Peter’s similarity, for example, is identified with
Peter’s whiteness as respecting John’s whiteness and John’s similarity
with John’s whiteness as respecting Peter’s whiteness. This way the
traditional picture of relations as accidents inhering in single subjects
can be preserved even on the reductionist picture.
71. DM XLVII.1.25: “…cum dicitur denominatio respectiva consurgere ex coexistentia
plurium absolutorum absque ulla reali additione, non esse intelligendum illam de-
nominationem aeque ac simul sumi ex pluribus formis absolutis, una intrinseca, et
altera extrinseca. Sed intelligendum est, illam denominationem requirere quidem con-
sortium, seu coexistentiam talium rerum seu formarum: tamen in unoquoque extremo
sumi a propria forma ut respicit aliam, quae ut sic habet rationem relationis, quamvis
in re non sit alia ab ipsa forma absoluta.” My thanks to Cecilia Trifogli for draw-
ing my attention to the significance of this passage.
philosophers’ imprint –  14  – vol. 13, no. 2 (january 2013)
3.2 Non-mutual relations
Even more important for Suárez than the issue of relations as single-
subject accidents is the possibility of non-mutual relations. In addition
to the warning at the end of section 2, he writes one of the longest
sections of On Real Relation in order to defend their possibility.72 Non-
mutual relations should not be confused with asymmetrical relations,
such as Peter’s being taller than John. The converse relation to Peter’s
being taller is John’s being shorter, which is a different species of
relation but, nevertheless, is an equally real relation. Non-mutual
relations are cases where one relation is real and its converse is not
real.73 The notion may sound strange to modern ears, but the existence
of non-mutual relations is widely accepted by medieval philosophers.
It is easy to see how one might accommodate non-mutual relations
on a non-reductionist picture (though seeing this may not remove all
strangeness). The non-reductionist says that in order for a subject to
change from not being related to something else to being really related,
some new thing or mode must be added to the subject. In the case of
mutual relations, both relata receive such a new thing or mode. In the
case of non-mutual relations, only one of the relata receives such a
new thing or mode. Hence, the relatum that received the new entity
will be the subject of a new real relation, while the other relatum will
merely be the subject of a relation of reason that our minds imagine as
the converse of the real relation.74
This account is not available to the reductionist. Presumably
Gregory of Rimini and the other nominalists whom Suárez cites as
denying that there are any non-mutual relations recognized that
72. DM XLVII.2.25 and XLVII.15.
73. DM XLVII.15.1.
74. For a few more details on how one non-reductivist handles non-mutual rela-
tions, see Giorgio Pini, “Can God Create My Thoughts? Scotus’s Case against
the Causal Account of Intentionality”, Journal of the History of Philosophy 49
(2011): 56–58.
 sydney penner Suárez on the Reduction of Categorical Relations
reductionism does not lend itself to non-mutual relations and so
decided to jettison them.75 Suárez, however, wants to preserve them.
There are two groups of cases of particular concern to him. The first
group comprises relations between created things and God. It was a
standard theological teaching that creatures bear real relations to God
but that God is not the subject of any real relations to creatures. So,
for instance, Peter’s being created is a real relation, but God’s creating
him is not.
The second group of concern to Suárez requires some stage-setting.
Aristotle divided relations into three classes. Interpretations of just how
this division goes vary, but we might say that there are those founded
on unity or multitude, those founded on potencies of acting or suffering,
and those founded on measure and the measurable.76 Suárez devotes
several sections of On Real Relation to defending Aristotle’s division,77
even though his defence in the end looks rather deflationary relative
to the more wholehearted defences offered by Aquinas and others.78
Aristotle remarks that a distinctive feature of the third class is that for
each relative that is truly said to be towards something, the terminus
in turn “is called relative because something else is related to it” rather
than because it truly is towards the first relatum.79 For example, the
knowable is called knowable merely because knowledge is related to it.
This claim in Aristotle initiated a tradition of seeing non-mutuality as a
distinguishing feature of the third class. Suárez, as might be expected,
is loath to abandon the tradition.
Perhaps he should not have been so loath. The section in which
Suárez defends non-mutual relations is one of the less satisfying of the
treatise. The arguments he gives on behalf of his opponents seem more
compelling than the ones he endorses, some of his own arguments
75. DM XLVII.15.7.
76. Aristotle makes the division in Metaphysics V.15 1020b26–1021b3. Suárez in-
troduces it in DM XLVII.10.1–4.
77. DM XLVII.10–14.
78. See especially DM XLVII.10.15–16.
79. Metaphysics V.15 1021a29–30.
philosophers’ imprint –  15  – vol. 13, no. 2 (january 2013)
seem obviously unworkable, and in places he says things that suggest
he in fact recognizes that he has little reason to advocate non-mutual
relations. The section is long and complex, and I will not analyze all its
arguments in detail, but I will present some representative passages to
support my judgements.
Recall that non-mutuality is supposed to be distinctive of the third
class of relations in Aristotle’s division — e. g., psychological relations
such as those between knowledge and the knowable and between
sight and the visible. Objections can come from two directions. They
can target the claim that relations of the third class are non-mutual, or
they can target the claim that relations of the first two classes are always
mutual (if some relations in the first two classes are non-mutual, then
non-mutuality will not be distinctive of the third class). The second
kind of objection is especially pressing for philosophers who wish to
maintain the theological doctrine that God is not the subject of any
real relations, since many of the relations between God and creatures
do not seem to fit into the third class.
Suárez considers several arguments of each sort. The first argument
for the mutuality of relations in the third class starts from the premise
that the most promising cases of non-mutuality are the relations
between knowledge and the knowable and between sensation and
the sensible. But these relations are in fact mutual, the argument goes,
and so we have no reason to think that there are any non-mutual
relations. The argument for the premise that these relations are mutual
is that the denomination ‘knowable’, for instance, is relative rather
than absolute. That is, the knowable is so in relation to knowledge,
just as knowledge is so in relation to the knowable. Furthermore, that
something is a knowable is grounded in reality, in things themselves.
In other words, knowability is not merely a relation we imagine. So
a knowable’s relation to knowledge is real. Everyone grants that
knowledge’s relation to the knowable is real. Since both relations are
real, we have mutuality.80
80. DM XLVII.15.3.
 sydney penner Suárez on the Reduction of Categorical Relations

In response, Suárez makes a distinction: Of course, presumably there is a reason why Aristotle and so many
philosophers after him thought that relations of the third class are
Therefore, in response to the first argument made at
non-mutual. And in fact there does seem to be something importantly
the beginning, it should be said that the knowable can
asymmetrical about these psychological relations. We might start with
be denominated in two ways: first, merely terminatively
the observation that the real change is always on one side (knowledge,
and, as it were, passively, and second, correlatively with
sensation) and the Cambridge change always on the other (the
knowledge. In the first way it is denominated extrinsically
knowable, the sensible). More fundamentally, the psychological states
from the real knowledge itself and in this way such a
have an aboutness or directedness that their objects do not have, a
denomination can be said to be in things themselves
directedness that might readily be thought of as a relational sort
rather than constructed through the intellect…. The
of thing if ever there was one. Hence, it is not surprising that non-
second denomination of the knowable is respective and
reductive realists think there is some real, new relational entity on the
this is only through a relation of reason, since our minds,
one side but not the other. Since ― on the non-reductionist view ―
in order to explain the relation that knowledge has to that
real relations require such a new entity in a subject for it to be really
knowable, conceive that [knowable] as correlative to the
related to something else, the case as described will be one of a non-
knowledge. The argument, therefore, only goes through
mutual relation.
concerning the former denomination.81
But Suárez is a reductionist. For a reductionist, a mere Cambridge
change is sufficient for something to become the subject of a real
Making a distinction in the face of a difficulty is, of course, the
relation. No additional entity is required in the subject. So the fact
strategy perfected by scholastic philosophers. And, indeed, in many
that there is no real change on the side of the knowable or the sensible
cases a distinction is just the cure needed. In this case, however, the
is no reason not to ascribe real relations to them. To put the point
distinction seems both obscure and ad hoc. It is difficult to shake
in a different way, it may be true that there is something interesting
the impression that it is a distinction made in a desperate attempt
happening on the side of knowledge that is not happening on the side
to shore up a thesis held primarily out of respect for the authority
of the knowable. But this interesting thing is also not happening in
of philosophical predecessors, or, as Suárez puts it, “a view that is
the case of a white object that becomes subject to a similarity relation
received with so much consensus that philosophers are not permitted
because some other object turned white (the whiteness does not
to call it into doubt”.82
have that special aboutness or directedness that intentional states
81. DM XLVII.15.13: “Ad primum ergo argumentum in principio positum dicendum est,
have). So the right conclusion to draw is that there is something extra
scibile dupliciter posse denominari: primo mere terminative et quasi passive: secundo going on with knowledge and its kin, rather than that the knowable
correlative ad scientiam. Primo modo denominatur extrinsece ab ipsamet scientia fails to be a real relation. Whatever the similar white object has, the
reali, atque ita talis denominatio dici potest esse in rebus ipsis, et non esse per intel-
lectum conficta…. Secunda denominatio scibilis est respectiva, et haec solum est per knowable also has.
relationem rationis, quia mens nostra ad explicandam illam relationem quam scientia The main argument of the second sort — i. e., against the claim that
habet ad ipsum scibile, concipit illud ut correlativum scientiae. Argumentum ergo so-
lum procedit de priori denominatione.” only relations of the third class are non-mutual — rests on the premises
82. DM XLVII.15.8: “Estque sententia tanta consensione recepta, ut non liceat philoso- that many of the relations between creatures and God belong to the
phis eam in dubium revocare.” first and second classes and that relations between creatures and God

philosophers’ imprint –  16  – vol. 13, no. 2 (january 2013)

 sydney penner Suárez on the Reduction of Categorical Relations
are non-mutual. From these premises it follows that at least some
relations in the first and second classes are also non-mutual.83 Suárez
basically concedes, granting that there are non-mutual relations in the
first and second classes. He tries to preserve the force of the Aristotelian
claim, however, by arguing that all relations in the third class are non-
mutual by virtue of being the sort of relations that belong to that class.
Relations in the first and second class only rarely fail to be mutual, and
when they do fail, this is not because of something about the relation
itself but rather because of features of the relata.84
In particular, Suárez now posits another necessary condition for
real relations: the two relata must belong to the same order.85
A real relation consists in an order of one thing to another;
it is, therefore, rightly required for such a relation that it
be between extremes of the same order.86
The relevant orders turn out to be created being and uncreated
being. Hence, God, the uncreated being, is not the subject of any real
relations to creatures.87 It is difficult to see why an opponent should be
moved by Suárez’s claims here. Why not just dismiss Suárez’s claims
as ad hoc and otherwise entirely unmotivated? But there is an even
more serious problem. Suárez is trying to defend non-mutual relations
between creatures and God, but this condition that the relata be of
the same order does not achieve that. This condition merely has the
consequence that the relations between creatures and God fail to
be real in each direction — i. e., the relations are relations of reason
in both directions. The obvious solution, of course, is to revise the
83. DM XLVII.15.6.
84. DM XLVII.15.22–29.
85. Cf. Thomas M. Ward, “Relations Without Forms: Some Consequences of
Aquinas’s Metaphysics of Relations”, Vivarium 48 (2010): 292.
86. DM XLVII.15.24: “…relatio realis consistit in ordine unius rei ad aliam: merito ergo
ad talem relationem requiritur ut sit inter extrema eiusdem ordinis.”
87. DM XLVII.15.23.
philosophers’ imprint –  17  – vol. 13, no. 2 (january 2013)
condition to say that real relations require that the terminus not be of
an inferior order.88 But that only worsens the worry that this is an ad
hoc condition.
Part of what makes Suárez’s account look odd is that he appears
to recognize that he does not need to make it: that he does not need
to defend non-mutual relations. He goes out of his way to argue
that while it would indeed be both gravely erroneous and absurd to
say that the relations between creatures and God are mutual on the
non-reductionist picture, it is of little consequence to say so on the
reductionist picture. In fact, he suggests that we might be left with
a mere verbal disagreement.89 He puts his finger on just the right
point. The motivation for the theological doctrine comes from more
general doctrines about God’s immutability and simplicity. On a non-
reductionist picture, saying that God comes to be the subject of a real
relation amounts to saying that some new thing or mode, distinct from
his substance, comes to inhere in or modify him. God would then have
really changed (as opposed to merely Cambridge-changed) and would
be a composite, violating the doctrines of immutability and simplicity.
But none of this is a worry on the reductionist account, since, in this
account, becoming the subject of a real relation does not require the
addition of some new thing or mode in the subject.90
Suárez explicitly recognizes all this but then concludes:
But I do not, nevertheless, think that one should depart
from the way of speaking of St. Thomas and the more
weighty theologians.91
88. Cf. DM XLVII.15.27: “In God, nevertheless, that relation is non-mutual on ac-
count of the eminence of his entity” (“…tamen in Deo, propter eminentiam suae
entitatis, relatio illa est non mutua …”).
89. DM XLVII.15.20.
90. DM XLVII.15.17–21. For more on this issue in the context of Aquinas, see
Ward, “Relations Without Forms”, 279–301.
91. DM XLVII.15.22: “Nihilominus tamen non censeo esse recendum a modo loquendi D.
Thomae, et graviorum Theologorum …”
 sydney penner Suárez on the Reduction of Categorical Relations
But if rejecting non-mutual relations on a reductionist account is of
little consequence, why should one not depart? This seems to be a case
where his respect for the great philosophers of the past ― admirable as
such respect may often be ― led him astray. Fundamentally, Suárez has
accepted a reductionist account; trying to clothe it in non-reductionist
“ways of speaking” makes for an ill fit.
4.  Different kinds of reductionism
By Suárez’s own telling, his account “hardly differs” from the standard
nominalist account. There may be disagreement about just how to
analyze relations and about what sorts of relations to posit in particular
cases, but both accounts are thoroughly reductionist about categorical
relations. At the basic ontological level, we need not posit any new
entities to account for categorical relations. Two substances bearing
the accident whiteness are ipso facto similar to each other. No further
peculiar relational accident or mode is required.
Not only does Suárez identify relations with absolute accidents in
the initial presentation of his own view,92 but he repeatedly emphasizes
throughout the remainder of the treatise that, on his view, relations
require nothing in addition to the foundation and terminus. Once the
foundation and terminus are posited, the relation follows ipso facto. In
fact, this claim becomes a premise in other arguments. For example,
in response to worries about an infinite number of relations in certain
cases (e. g., an infinite number of equality and inequality relations in
a continuum93), Suárez notes that there really is no problem with an
infinity of relations, since they do not add anything real anyway.94 So,
with respect to this core of the reductionist account, Suárez is in perfect
agreement with Ockham, Gregory of Rimini, and other nominalists.
92. DM XLVII.2.22.
93. DM XLVII.9.5.
94. Here is a partial list of the places where Suárez says that relations follow once
the foundation and terminus are posited and/or that categorical relations add
nothing new: DM XLVII.2.24, 4.12, 4.13, 4.20, 5.12, 6.1, 7.1, 7.4, 8.7, 8.8, 8.14, 9.5,
11.12, and 11.19.
philosophers’ imprint –  18  – vol. 13, no. 2 (january 2013)
There is, however, a distinctive element to Suárez’s account,
because of his commitment to relations as one-subject accidents and
to non-mutual relations. We can get a clearer sense of the distinctive
element by looking at the criticisms made of Suárez’s account by
the Bohemian scholastic Rodrigo de Arriaga (1592–1667).95 Arriaga
does not share Suárez’s commitment to non-mutual relations and to
relations as one-subject accidents, so he straightforwardly reduces a
relation to its foundation and terminus, both of which he thinks are
parts of the relation. His preferred analogy for a relation is being
two in number. Just as twoness is not needed as some third thing in
addition to the two individual objects in order for it to be true that
there are two objects, so similarity is not needed as some third thing
in addition to two white objects in order for it to be true that they
are similar to each other. As we saw, however, Suárez would object
that, on Arriaga’s account, the foundation and terminus contribute
in the same way to the relation, a charge that Arriaga would happily
grant. Since Suárez wishes to avoid that result, he identifies a relation
with its foundation and only its foundation. To be sure, he talks of
the foundation as respecting its terminus. But, formally speaking, a
relation is identical to its foundation on his view.
Arriaga takes Suárez to task for this identification. It seems as if
Suárez has not heeded the force of the Separability Argument, which
shows that the foundation of a relation can continue to exist while the
relation passes away. Both Arriaga and Suárez reasonably respond that
the Separability Argument does not show that a relation is separable
from the foundation and terminus taken together, and so the argument
does not establish that reductionism cannot work. But surely the
argument does successfully show that a relation is separable from the
foundation alone. Yet Suárez has to deny that, since he identifies the
relation with its foundation and identical things cannot be separated.
95. Arriaga addresses Suárez’s account in Cursus philosophicus (Antwerp:
Balthasar Moreti, 1632), 169–72. For more on Arriaga’s own account and his
criticisms of Suárez, see my “Rodrigo de Arriaga on Relations”, The Modern
Schoolman 89 (2012): 25–46.
 sydney penner Suárez on the Reduction of Categorical Relations
So if John, for example, receives a dark tan, on Suárez’s account the
relation of similarity in Peter ― being formally identical with his
whiteness ― still exists “fundamentally or in proximate aptitude”.96
Granted, Suárez would deny that it is still appropriate to denominate
Peter similar to John, since the denomination also connotes John
and his whiteness, which no longer exists. But Arriaga has reason to
protest that it is a strange result if Peter’s similarity to John can, in any
way whatever, survive John’s becoming dark.
Insofar as Suárez’s motivations for identifying a relation with its
foundation alone are underwhelming, one would reasonably avoid
this strange result.
5.  Does reductionism work?
A key philosophical question is, of course, whether any kind of
reductionism about relations works. While I will not attempt a
thorough treatment of that question here, some comments are in order.
5.1 Irreducible relational statements
Reductionism might be deemed a nonstarter thanks to twentieth-
century logical results that allegedly prove that relational statements
cannot be reduced to non-relational statements. And, indeed, putative
reductions tend to leave suspiciously relational-looking predicates in
their reductive bases. But note that Suárez is interested in an ontological
reduction. That is, he claims that we do not need to posit any irreducible
items among the world’s furniture to account for categorical relations.
He is not claiming that relational statements can likewise be reduced.
Some philosophers may be committed to an isomorphism between
reality and language such that ontological reductionism requires
linguistic reductionism. Perhaps Leibniz is an example.97 But Suárez is
96. DM XLVII.2.23.
97. See Jauernig, “Disentangling Leibniz’s Views on Relations and Extrinsic De-
nominations”, 198–204.
philosophers’ imprint –  19  – vol. 13, no. 2 (january 2013)
not. So the quick answer to objections of this sort is that the objections
are irrelevant to Suárez’s project.98
5.2 External Relations
The case of a similarity relation between two white things ― a favourite
example in medieval philosophy ― lends itself to reductionist
treatment. Other cases, however, less obviously do. These other cases
might give rise to scepticism about a general reductionist programme.
The standard division of relations in contemporary discussions will
be helpful here. The division is simple. Relations come in two flavours:
internal and external. We can characterize them as follows:
• X and Y stand in an internal relation iff the relation
necessarily follows from X’s and Y’s intrinsic properties.
• X and Y stand in an external relation iff X and Y stand in a
relation that is not an internal relation.99
The idea here is straightforward. An example of an internal relation
would be the similarity between Peter and John. The intrinsic monadic
properties of whiteness in each of them necessitate the relation of
similarity. But take the spatial relation of being five metres apart that I
mentioned earlier. Here it does not seem as if the intrinsic properties
will suffice for the relation. We could move Peter another five metres
from John. Peter and John seem still to have all the absolute properties
that they had before Peter was moved, but they no longer have the
relation of being five metres apart. Now they have the relation of
being ten metres apart. To use Suárez’s terminology, this relation
98. Jeffrey E. Brower has a sensible, longer discussion of the viability of onto-
logical reductionism about relations in the face of the logical results. See
“Abelard’s Theory of Relations”, 605–31.
99. Both David Lewis (On the Plurality of Worlds [Malden, MA: Blackwell, 1986],
62) and D. M. Armstrong (Universals, 43–44) make this distinction. Lewis
goes on to make a further distinction among those that I am calling external
relations and retains the term for one of the resulting subcategories. We can
ignore that complication.
 sydney penner Suárez on the Reduction of Categorical Relations
seems not to follow necessarily once the foundation and terminus
have been posited.
Medieval philosophers were aware of the distinction between
internal and external relations. Scotus is well known for making
the distinction under the terms ‘intrinsically occurring relations’
(‘relationes intrinsecus advenientes’) and ‘extrinsically occurring relations’
(‘relationes extrinsecus advenientes’).100 Some philosophers made this
distinction a prominent part of their theories of relations. Durandus
of St. Pourçain, for example, seems to have recognized the apparent
ontological implications and given a reductionist account of internal
relations, such as similarity, but a non-reductionist account of external
relations, such as spatial relations.101 In general, however, medieval
philosophers make less of this distinction than one might expect. It
is as if they assume that relations require, or at least can receive, a
unified treatment, rather than that this distinction might point to the
need for different accounts for each type of relation.
Suárez is aware of the distinction. He only gives a cursory mention
to it in On Real Relation,102 but he discusses it at more length in his
treatise on action.103 Moreover, there are several places where he
recognizes cases of putatively external relations as problematic. His
responses suggest that he will take different tacks depending on the
case but that he will insist that categorical relations can be reduced
in all cases.
100. S
 cotus was well known for making the distinction, but there was signifi-
cant disagreement about precisely which distinction he meant to capture
(Suárez enters the fray in DM XLVIII.1.2–8). The different options corre-
spond to some of the different notions that go by the term ‘internal rela-
tion’ in twentieth-century philosophy. I defined the term above according
to its most common usage recently, but that is certainly not the only way
the term is used. See Schaffer, “The Internal Relatedness of All Things”.
101.  ee Suárez’s presentation of Durandus’s view in DM XLVII.2.4–6. See also
S
Elizabeth Lowe, The Contested Theological Authority of Thomas Aquinas: The
Controversies between Hervaeus Natalis and Durandus of St. Pourçain (New
York: Routledge, 2003), 89 ff.
102. DM XLVII.4.12.
103. DM XLVIII.1.2–8.
philosophers’ imprint –  20  – vol. 13, no. 2 (january 2013)
Suárez considers spatial relations when responding to Durandus’s
position. Suárez simply denies that spatial relations are external. Peter’s
similarity to John does not necessarily follow if we just posit Peter and
John irrespective of colour. Rather, it follows necessarily if we posit
Peter and John and their respective whitenesses. Of course, that is
perfectly in keeping with our characterization of internal relations. But
Suárez thinks that if we are tempted to see spatial relations as external,
we are making a mistake akin to forgetting to put Peter’s whiteness and
John’s whiteness in the reduction base. Once we posit Peter and John
along with their respective places, the relevant spatial relations follow
necessarily in just the way that similarity relations follow necessarily.104
The obvious point to press here is about the nature of places.
What is Peter’s place? Suárez does not address the issue here,105 but
either place is an absolute feature of Peter or it is a relative feature.
That Peter’s place is an absolute or intrinsic feature — i. e., a feature
that does not depend on any other things in the world — will seem
deeply implausible to many.106 But the second option leaves the
reduction looking unsuccessful. What good is it to reduce a relation
to another relation?
But here we should remember that Suárez thinks there are both
categorical and transcendental relations and that he is only attempting
a reduction of categorical relations here. So his reduction can still be
successful even though not everything relational has been eliminated
from the reduction base, provided that all the relations that remain are
transcendental relations.
This assumes, of course, that Suárez is not trying to reduce
categorical relations solely to absolute things. One might well think
104. DM XLVII.2.5.
105.  uárez does, of course, have a treatise on the Aristotelian category of where,
S
in which he says that it is an absolute mode distinct ex natura rei from the
subject, though, thanks to our cognitive limitations, we can understand it
only relationally — i. e., as the foundation of relations of nearness and dis-
tance (DM LI.1.13–14.). But where is only a relation according to being said.
106. B
 ut see Keith Campbell, Abstract Particulars (Oxford: Blackwell, 1990),
126–30.
 sydney penner Suárez on the Reduction of Categorical Relations
that he is trying precisely such a reduction. In most of his discussion,
the term ‘absolute form’ is used frequently; one certainly gets the
impression that absolute things are what Suárez has in mind when
he thinks about what to reduce categorical relations to. Furthermore,
he provides no explicit qualification when presenting his reductionist
view that would let readers know to expect relations, albeit non-
categorical ones, in the reduction base.
In a later section of the treatise, however, Suárez considers the
paternity relation and clearly lets transcendental relations in the door,
whether consciously or not. Suppose Peter is the father of Paul — i. e.,
is the subject of a paternity relation to Paul. Suárez considers the
scenario in which God creates duplicates of Peter and Paul with
all their absolute properties. This is just the right sort of thought
experiment to determine whether a relation is internal or external.
Suárez recognizes the difficulty and concedes that the duplicate Peter
would not be the father of the duplicate Paul. But he thinks he has a
solution: posit a third thing in the reduction base. On this account,
the paternity relation requires Peter, Paul, and an action of generation.
Once the action has been added, then, Suárez says, the paternity
relation follows necessarily.107
But actions are certainly not wholly absolute things in Suárez’s
view. As he makes clear in the first section of his treatise on action,
an action essentially includes transcendental relations.108 So, in the
case of paternity, at least, Suárez leaves transcendental relations
in the reduction base. It is difficult to tell just what to make of this
development. Is Suárez, when confronted with a problematic case,
conceding something that undermines his official reductionist
position? Or is this continuous with the rest of what Suárez says?
Leaving some relations in the reduction base is certainly less
ambitious than reducing relations to only absolute things. Some
particular reductions — e. g., in the case of similarity — would be
107. DM XLVII.7.12.
108. DM XLVIII.1–2.
philosophers’ imprint –  21  – vol. 13, no. 2 (january 2013)
reducible to wholly absolute entities, but others would not be. Leaving
irreducible transcendental relations in one’s ontology is a significant
concession to the non-reductionists. Of course, insofar as external
relations pose a serious challenge to thoroughgoing reductionism
about relations, this concession might be a gain in plausibility. But it is
not clear, unfortunately, whether Suárez actually intends to make this
concession to non-reductionism or whether he conceded something
without noting its implications.109
6.  Relations as a category
There is a leftover item from earlier in this paper (§2.1) that still needs
resolution. We looked at two arguments for the claim that relations
are not real beings and that therefore relations also fail to form a
distinct genus of real beings: the Cambridge Change Argument and
the Necessary Consequence Argument. A reductionist might well be
sympathetic to these arguments, but Suárez, recall, concludes that
relations do form a distinct category of being. Suárez’s response to the
arguments is illuminating. He points out the assumption that relations
have real being only if they are really distinct from absolute beings.
Suárez denies the assumption. So he agrees that relations add no new
being (to this extent he is in sympathy with the arguments), but he
does not think that it follows that relations are not real. As he puts it:
Hence, everything brought forward for the proof in that
third argument is useful for showing how a real relation
does not add any thing or real mode to absolute things
that is ex natura rei distinct from the absolute things, but
they are not at all valid for concluding that a relation of
this sort is wholly nothing.110
109. M
 ight Suárez think that he can in turn reduce transcendental relations to
absolute things? Perhaps, but I see no evidence for such a programme in
the text.
110. 
DM XLVII.2.24: “Unde probationes omnes, quae in illo tertio argumento afferun-
tur, utiles sunt ad declarandum, quomodo relatio realis non addat rebus absolutis
 sydney penner Suárez on the Reduction of Categorical Relations

Instead of eliminating relations, we should reduce them, and reduced Bibliography

relations are still real relations.
Adams, Marilyn McCord. “Relations, Inherence, and Subsistence: or,
Not only does Suárez think that categorical relations are real beings,
Was Ockham a Nestorian in Christology?”. Noûs 16 (1982): 62–75.
but he argues that they form one of the basic Aristotelian categories
———. William Ockham. 2 vols. Notre Dame, IN: University of Notre
of being. This might come as a surprise even to those who are happy
Dame Press, 1987.
to grant that reduced things are real things. We might have expected
Ariew, Roger. “Descartes and Leibniz as Readers of Suárez: Theory
the categories to be a classification of foundational being. But if, for
of Distinctions and Principle of Individuation”. In The Philosophy
instance, a relation of similarity is nothing more than the whiteness of
of Francisco Suárez, edited by Benjamin Hill and Henrik Lagerlund,
its foundation, why should the whiteness go in one category (quality)
38–53. Oxford: Oxford University Press, 2012.
and the similarity go in another (relation)? But Suárez says that a
Aristotle. The Complete Works of Aristotle: The Revised Oxford Translation.
distinction of reason with some foundation in reality is sufficient for a
Edited by Jonathan Barnes. 2 vols. Princeton, NJ: Princeton
distinction of categories. This is not unique to the category of relations;
University Press, 1984.
he thinks action and passion are also two categories divided by a
Armstrong, D.M. Universals: An Opinionated Introduction. Boulder, CO:
mere distinction of reason with some foundation in reality.111 What is
Westview Press, 1989.
at work here, however, is more a peculiar account of what makes a
Arriaga, Rodrigo de. Cursus philosophicus. Antwerp: Balthasar Moreti,
category than something distinctive about his account of relations;112
1632.
the former is a topic for another time.113
Augustine. De Trinitate. Corpus Christianorum, series latina, vols. 50 and
50A. Turnhout: Brepols, 1968. English translation by Edmund Hill
in The Works of Saint Augustine: A Translation for the 21st Century, vol.
1.12. Brooklyn, NY: New City Press, 1991.
Beretta, Beatrice. Ad aliquid: la relation chez Guillaume d’Occam. Fribourg:
Éditions Universitaires Fribourg Suisse, 1999.
rem, vel realem modum, ex natura rei distinctum ab ipsis, non tamen quicquam
valent ad concludendum huiusmodi relationem nihil omnino esse.” Brower, Jeffrey E. “Abelard’s Theory of Relations: Reductionism and the
Aristotelian Tradition”. The Review of Metaphysics 51 (1998): 605–31.
111. DM XLVII.2.22. Cf. 2.24.
———. “Medieval Theories of Relations”. In The Stanford Encyclopedia
112. William E. McMahon has a brief discussion of Suárez on the categories in
“The Categories in Some Post-medieval Spanish Philosophers”, in Medieval of Philosophy (Winter 2010 Edition), edited by Edward N. Zalta.
and Renaissance Logic in Spain: Acts of the 12th European Symposium on Medieval URL =  < http://plato.stanford.edu/archives/win2010/entries/
Logic and Semantics, Held at the University of Navarre (Pamplona, 26–30 May
1997), ed. Ignacio Angelelli and Paloma Pérez-Ilzarbe (Hildesheim: Georg relations-medieval/>.
Olms Verlag, 2000), 355–70. ———. “Relations Without Polyadic Properties: Albert the Great on the
113. My thanks to the members of a most enjoyable and illuminating reading Nature and Ontological Status of Relations”. Archiv für Geschichte der
group on DM 47 during Hilary and Trinity terms 2011, especially Cecilia Tri-
Philosophie 83 (2001): 225–57.
fogli, Daniel Cairns, and Nicolas Vaughan. Thanks, too, to Jeffrey E. Brower
and Scott MacDonald for helpful conversations, to Erin Penner for spotting Burns, J. Patout. “Action in Suárez”. The New Scholasticism 38 (1964):
numerous unhappy expressions in an earlier draft, and to two anonymous 453–72.
referees for this journal for their comments.

philosophers’ imprint –  22  – vol. 13, no. 2 (january 2013)

 sydney penner Suárez on the Reduction of Categorical Relations
Campbell, Keith. Abstract Particulars. Oxford: Blackwell, 1990.
Decorte, Jos. “‘Modus’ or ‘Res’: Scotus’ Criticism of Henry of Ghent’s
Conception of the Reality of a Real Relation”. In Via Scoti:
Methodologica ad mentem Joannis Duns Scoti, edited by Leonardo
Sileo, vol. 1, 407–29. Rome: Edizioni Antonianum, 1995.
Eschweiler, Karl. “Roderigo de Arriaga S. J.: Ein Beitrag zur Geschichte
der Barockscholastik”. Gesammelte Aufsätze zur Kulturgeschichte
Spaniens 3 (1931): 253–85.
Freddoso, Alfred J. Introduction to On Creation, Conservation, and
Concurrence: Metaphysical Disputations 20–22. South Bend, IN: St.
Augustine’s Press, 2002.
Glauser, Richard. “Descartes, Suarez, and the Theory of Distinctions”.
In The Philosophy of Marjorie Grene, edited by Randall E. Auxier and
Lewis Edwin Hahn, 417–45. Chicago: Open Court, 2002.
Gracia, Jorge J.E. “Francisco Suárez: The Man in History”. American
Catholic Philosophical Quarterly 65 (1991): 259–66.
Hansen, Heine. “Strange Finds, or Nicholas of Paris on Relations”. In
Logic and Language in the Middle Ages: A Volume in Honour of Sten
Ebbesen, edited by Jakob Leth Fink, Heine Hansen, and Ana María
Mora-Márquez, 139–154. Leiden: Brill, 2012.
Henninger, Mark G. Relations: Medieval Theories 1250–1325. Oxford:
Clarendon Press, 1989.
Jauernig, Anja. “Disentangling Leibniz’s Views on Relations and
Extrinsic Denominations”. Journal of the History of Philosophy 48
(2010): 171–205.
John Duns Scotus. Opera omnia, edited by C. Balić et al, vol. 7. Città del
Vaticano: Typis Polyglottis Vaticanis, 1973.
Lewis, David. On the Plurality of Worlds. Malden, MA: Blackwell, 1986.
Lowe, Elizabeth. The Contested Theological Authority of Thomas Aquinas:
The Controversies between Hervaeus Natalis and Durandus of St. Pourçain.
New York: Routledge, 2003.
Manser, G.M. “Begriff und Bedeutung der transzendentalen Beziehung”.
Divus Thomas 4 (1941): 351–60.
philosophers’ imprint –  23  – vol. 13, no. 2 (january 2013)
McMahon, William E. “The Categories in Some Post-medieval Spanish
Philosophers”. In Medieval and Renaissance Logic in Spain: Acts of the
12th European Symposium on Medieval Logic and Semantics, Held at the
University of Navarre (Pamplona, 26–30 May 1997), edited by Ignacio
Angelelli and Paloma Pérez-Ilzarbe, 355–70. Hildesheim: Georg
Olms Verlag, 2000.
Menn, Stephen. “Suárez, Nominalism and Modes”. In Hispanic
Philosophy in the Age of Discovery, edited by Kevin White, 226–56.
Washington, D.C.: Catholic University of America Press, 1997.
Mugnai, Massimo. Leibniz’ Theory of Relations. Stuttgart: Franz Steiner,
1992.
Murray, Michael V. “The Theory of Distinctions in the Metaphysics of
Francis Suarez”. Ph.D. dissertation, Fordham University, 1945.
Panaccio, Claude. “Connotative Terms in Ockham’s Mental Language”.
Cahiers d’épistémologie no. 9016. Montreal: Université du Québec à
Montréal, Group de Recherche en Épistémologie Comparée, 1990.
Penner, Sydney. “‘An Accident that is Simultaneously in Two Subjects’:
Leibniz and Some Predecessors on the Possibility of Two-Subject
Accidents”. Unpublished.
———. “Rodrigo de Arriaga on Relations”. The Modern Schoolman 89
(2012): 25–46.
Pini, Giorgio. “Can God Create My Thoughts? Scotus’s Case against the
Causal Account of Intentionality”. Journal of the History of Philosophy
49 (2011): 39–63.
Rubio, Antonio. Logica mexicana, hoc est commentarii breviores et maxime
percipue in universam Aristotelis dialecticam. Lyons: Antonio Pillehotte,
1620.
Russell, Bertrand. “Logical Atomism”. In Logic and Knowledge: Essays,
1901–1950, edited by Robert Charles Marsh, 321–44. London: Allen
and Unwin, 1956.
———. Principles of Mathematics. London: Routledge, 2010.
Schaffer, Jonathan. “The Internal Relatedness of All Things”. Mind 119
(2010): 341–76.
 sydney penner Suárez on the Reduction of Categorical Relations

Secada, Jorge. “Suárez on the Ontology of Relations”. In Interpreting

Suárez: Critical Essays, edited by Daniel Schwartz, 62–88. Cambridge:
Cambridge University Press, 2011.
Shields, Christopher. “Shadows of Beings: Francisco Suárez’s Entia
Rationis”. In The Philosophy of Francisco Suárez, edited by Benjamin Hill
and Henrik Lagerlund, 57–74. Oxford: Oxford University Press, 2012.
Soto, Domingo de. In Porphyrii Isagogen, Aristotelis Categorias, librosque
De demonstratione, absolutissima commentaria. Venice: apud Guerraeos
fratres et socios, 1573.
Spade, Paul Vincent. “Ockham, Adams and Connotation: A Critical
Notice of Marilyn Adams, William Ockham”. The Philosophical Review
99 (1990): 593–612.
———. “Ockham’s Distinctions Between Absolute and Connotative
Terms”. Vivarium 13 (1975): 55–76.
Suárez, Francisco. De Incarnatione Verbi. Vols. 17–19 of Opera omnia.
———. Disputationes metaphysicae. 2 vols. Salamanca: John and Andreas
Renaut, 1597.
———. On Real Relation (Disputatio Metaphysica XLVII): A Translation from
the Latin, with an Introduction and Notes. Translated by John P. Doyle.
Milwaukee: Marquette University Press, 2006.
———. Opera omnia. 26 vols. Paris: Vivès, 1856–78. Vols. 25 and 26
reprinted: Hildesheim: Georg Olms Verlag, 1965.
Thomas Aquinas. Summa contra Gentiles. 5 vols. Translated by Anton C.
Pegis et al. Notre Dame, IN: University of Notre Dame Press, 1975.
Tweedale, Martin. “Ockham’s Supposed Elimination of Connotative
Terms and His Ontological Parsimony”. Dialogue 31 (1992): 431–44.
Ward, Thomas M. “Relations Without Forms: Some Consequences of
Aquinas’s Metaphysics of Relations”. Vivarium 48 (2010): 279–301.
William of Ockham. Ockham’s Theory of Terms: Part I of the Summa Logicae.
Translated by Michael J. Loux. Notre Dame, IN: University of Notre
Dame Press, 1974.
Williams, Scott. “Henry of Ghent on Real Relations and the Trinity:
The Case for Numerical Sameness Without Identity”. Forthcoming
in Recherches de Théologie et Philosophie Médiévales.

philosophers’ imprint –  24  – vol. 13, no. 2 (january 2013)